Shopping engine history and nomenclature

Comparison shopping has been available on the Internet since 1995

While some of the major shopping engines have been around for almost 10 years, the first shopping engines emerged in the mid 1990s. These shopping engines were originally labeled " price bots" or "price comparison engines" in that they enabled consumers to find the best price on a book or compact disc. As opposed to most of today's shopping engines which are driven by data feeds, these first sites were "intelligent agents" which scoured particular sites for product information and prices.

BargainFinder is arguably the first consumer facing shopping engine. Bruce Krulwich was working for Arthur Andersen when he launched BargainFinder in 1995. Krulwich's business credentials were supplemented by a formidable educational background, with degrees from Carnegie Mellon (BS in Applied Math and Computer Science), Yale University (MS in Computer Science and Artificial Intelligence) and Northwestern University (PhD in Computer Science/Artificial Intelligence). BargainFinder was an Artificial Intelligence Shopping Agent, better known as a "price bot." The point of BargainFinder was to find the best price for a particular product. Early work focused on finding the best price on CDs from stores like CDNow.com.

Meanwhile, Oren Etzioni, who received his B.S. in computer science from Harvard University and his MS and Ph.D. in computer science from Carnegie Mellon University, was hard at work on NetBot in 1995-96 while working as a Professor of Computer Science at the University of Washington. Jango was the first commercial product launched from NetBot and it combined a couple of projects (MetaCrawler and ShopBot) that were developed by the University of Washington team. Netbot was purchased by Excite in 1997 for $35,000,000. This was the first in a long string of successful exits for shopping engines. Oren is currently a professor at the University of Washington, a Venture Partner with Madrona and was the founder of Farecast, which sold to Microsoft in 2007.

The following articles provide more information about BargainFinder and Netbot:

Information and Integration Agents, Bruce Krulwich, 1996
The Ideal Cost of Ignorance, Stevan Alberty, 1996
A Scalable Comparison Shopping Agent for the World Wide Web, Robert Doorenbos, Oren Etzioni, and Daniel S. Weld, 1997
Netbot Shops the Internet's Myriad Malls, Sharon Baker, 1997
Netbot, Inc. Debuts First Product - Jango - at PC Forum, 1997

Shopping Engine? Comparison Shopping Engine? Shopping Comparison Engine? Price Comparison Engine? What's the difference?

The traditional shopping engine lists one picture, one title, one description and all of the merchants that sell that particular product. Via this format, shopping engines are known for normalizing or "SKUing up" product listings to help consumers quickly and easily make apples-to-apples comparisons to make smart buying decisions. Therefore, the names "price comparison engine," "comparison shopping engine (CSE)," or "shopping comparison engine" seem to describe these sites well.

In the early days, crawlers and intelligent agents focused on consumer electronics as well as books, music and videos because the products in these categories came with uniquely identifiable information like Universal Product Codes (UPCs) and International Standard Book Numbers (ISBNs). These hard goods were easy to SKU up and normalize.

However, since 2005 online sales of softer goods such as clothing, home appliances, jewelry, furniture, sporting goods, and toys have gained momentum. While some products such as a Cuisinart blender might be somewhat easy to normalize, other items like sweaters and furniture can be much more difficult. In the case of a sweater, as opposed to seeing one title, description, and picture of a product (as you would for a MP3 player), what becomes more apparent are the filtering capabilities on product attributes. When searching for a red sweater, a shopper will see myraid offers from many merchants rather than a simple set of normalized listings. Consumers will then sort by price, pictures, ratings, reviews, and attributes such as size, material, and brand in order to find the sweater they prefer.

With soft goods so important and less products normalized/SKUd up, the terms "comparison engines" or "shopping comparison engines" are less relevant. Simply put, these sites are shopping search engines.